Characterizing bacterial infections rapidly is key to effective treatment and identifying the spread of antibiotic resistance e.g. extended spectrum beta-lactamase (ESBL). Current testing of human specimens primarily relies on culture, which requires infrastructure unavailable at many clinics, especially in low/middle income countries, and is too slow inform initial treatment. This is especially critical in neonatal infections (e.g. urinary tract infections, meningitis), where delays in effective treatment can result in lifelong disability or death. Due to this risk, powerful, broad spectrum antibiotics (e.g. 3rd generation cephalosporins) are empirically given. However, ESBL, which account for roughly 1/3 of the gram negative infections in SE Asia, can destroy cephalosporin antibiotics, leading to initial treatment failure. To address this, we have developed a rapid, mobile platform which can detect pathogen directly from urine or cerebrospinal fluid (CSF) and identify beta- lactamase related antibiotic susceptibility in < 20 mins. The platform uses disposable microfluidic cartridges into which human samples are injected. A dark probe which becomes fluorescent in the presence of ESBL is then introduced. Any change in fluorescence is detected by a smart phone camera which has been modified with additional clip-on optics. By repeating the analysis in the presence of an antibiotic, information on the antibiotic susceptibility can also be obtained. The end results of this diagnostic are presented on the smart phone screen for facile point of care use. The mobile phone aspect also allows de-identified epidemiological information to be automatically collected, which is of use in rapid infection control of outbreaks and general monitoring of regional antibiotic susceptibility. This study proposes to adapt and test the platform for use in Thailand. The first aim (months 0-6) modifies the U.S. design for sustainable use in Thailand (e.g. sourcing local materials). The second aim (months 6-12) calibrates the platform against pathogens that have already been characterized by standard microbiological methods. This is done by refrigerating a small amount of human sample (few hundred microliters of urine or CSF), which will be analyzed by the platform if standard lab results find bacteria. These data (~100 samples) will be used to tune the analysis against the local pathogen spectrum. The final aim (months 12-24) will validate the platform by directly testing patient samples in parallel with the standard microbiological lab analysis. The samples will be pre-selected based on rapid gram staining to ensure a high number of bacterial samples for statistical significance (~400 specimens). Knowledge of microfluidic cartridge construction and smart phone optics fabrication will also be transferred to Thailand to allow sustainable, long term use of the platform.